Excitotoxicity mediated through over-activation of N-methyl-D-aspartate (NMDA) receptors has been suggested to play role in several neurological disorders including Alzheimer's disease (AD). However, glutamate mediates excitatory neurotransmission throughout cortex whereas AD brain shows region-specific pathological changes. This localized excitotoxicity may arise from the variable expression of NMDA receptors that are highly responsive to stimulation, leading to excess calcium influx and neuronal cell death. In this project, possible NMDA receptor combinations that occur in particular regions of the AD brain we identified using conantokins.
The pharmacology of the NMDA receptor was examined in six pathologically affected and two relatively spared regions of cerebral cortex using tissue obtained at autopsy from AD cases and matched controls. Neither the affinity nor the density of [3H]MK-801 binding varied significantly between AD cases and controls in any area, indicating that the NMDA receptor present in these regions were qualitatively similar with respect to [3H]MK-801 binding. However, maximal enhancement induced by either glutamate or spermine was regionally variable, with glutamate-mediated maximal enhancement higher in control than in AD cases in pathologically spared regions, whereas spermine-mediated maximal enhancement was higher in controls in areas susceptible to pathological damage. This latter result potentially indicates that neurons highly responsive to spermine may be lost in AD.
Conantokin-G (con-G) is a small helical peptide isolated from the venom of the cone snail Conus geographus that acts as an NMDA receptor antagonist. Studies comparing two analogues of con-G revealed that Ala(7)-con-G had higher affinity than Lys(7)-con- G, defining two distinct binding sites in control brain. However, the affinity for Lys(7)- con-G was higher in AD brain than in control brain, whereas the reverse was true for Ala(7)-con-G. These data suggest that the subunit composition of NMDA receptors may be locally variable, and may change, in AD.
In situ hybridization and PCR studies in autopsy tissue can help define the receptor subtypes that occur in human brain, but do not yield functional information. To determine which NMDA receptor combinations underlie these regional and AD-associated changes, the pharmacology of different NMDA receptors was investigated in a mammalian cell expression system. Binding assays with selected ligands on transfected cells expressing a particular receptor subunit combination would allow the binding parameters to be compared directly with values obtained from different human brain regions. Unfortunately, the NMDA receptor clones available were limited and not optimal for mammalian expression. For example, the human NR1011, NR2A and NR2B were used for transfecting of COS and BHK cells but no antibiotics resistance gene was available for the selection of transfected cells. Several techniques were used to confirm successful transfections, such as co-transfection of a β-galactosidase expression vector, immunocytochemistry, and Western blotting. Although the transfections were successful, the binding assays did not give good signal to noise ratios probably because there was insufficient receptor density at the cell membranes.
Xenopus oocyte expression is an alternative to mammalian cell expression. Expression of NMDA receptors in Xenopus oocytes gives information about receptor function using electrophysiological approaches, in addition to affinities for selected ligands. Oocytes were injected with one NR1 subunit (NR1000, NR1010, NR1011, NR1100, NR1110, or NR1111), together with either NR2A or NR2B, to form functional receptors. Glutamate showed higher affinity for NR2A-containing receptors overall, regardless of the NR1 subunit present. Spermine (30 μM) decreased or increased glutamate affinity for the NMDA receptor depending on the subunits present. Conantokin-G is a noncompetitive inhibitor of polyamine responses but a competitive inhibitor of glutamate responses. Con- G and the more potent analogue Ala(7)-con-G showed subunit-specific inhibition, with strong inhibition at NR2B-containing receptors (Ki values 0.081-1.3 μM), while receptors containing the NR2A subunit showed > 10-fold decrease in the affinity for the peptide (Ki values 2.5-22.9 μM). The two peptides also showed NRl subunit-specific variations in potency. Ala(7)-con-G was generally equally or more potent than con-G except at NR1100/NR2A, where con-G was more potent. Both con-G and Ala(7)-con-G showed a subunit-specific action at NR2A-containing receptors, which depended on the NR1 subunit present. A biphasic concentration-response curve was obtained, with potentiation at low concentrations of the peptide and inhibition at higher concentrations. For Ala(7)-con-G, this biphasic effect was observed at both 3 μM and 30 μM glutamate for NR1011/NR2A and NR1010/NR2A, whereas it was only observed in the presence of 3 μM glutamate at NR1000/NR2A, NR1111/NR2A and NR1100/NR2A receptors. In addition, Ala(7)-con-G was solely inhibitory at NR1110/NR2A at both glutamate concentrations. Con-G showed a similar profile to Ala(7)-con-G at NR1011/NR2A and NR1111/NR2A receptors with potentiation at 3 μM glutamate. However, except for at NR1011/NR2A, con-G was inhibitory at all other receptor combinations in the presence of 30 μM glutamate.
The major findings in this study demonstrate regional variations in the pharmacology of the NMDA receptor that change in AD. Based on results obtained using an oocyte expression system, NR1011/NR2B combinations are most likely lost in AD. Particular subunit combinations may represent specific targets for drugs that are able to ameliorate AD. In addition to being useful tools for characterizing different NMDA receptor subunit combinations due to their subunit specificity, Conantokins or other molecules selective for NR1011/NR2B or other NMDA receptors lost in AD may have potential as protective agents against this type of excitotoxic damage.